Teams handling contracts or forms that need citations or redaction get bounding boxes and confidence scores, at double OCR 3's price.
Kenning's take
What it does
Mistral OCR 4.1 is the current version of the OCR 4 model Mistral announced in June 2026. Along with the extracted text it returns bounding boxes, labels for each block such as titles, tables or signatures, and confidence scores, which helps with citations, redaction and human checks. It is available through the Mistral API and Document AI, with self-hosting for enterprise customers.
- Paragraph-level bounding boxes
- Block labels such as titles, tables, equations and signatures
- Inline confidence scores
- 170 languages
- Single-container self-hosted deployment
- Batch processing
Who it suits. Developers and enterprises building search, RAG or document-processing pipelines that need layout as well as text.
Key facts
- Category
- AI models
- Part of
- Le Chat (Vibe by Mistral)
- Made by
- Mistral AI
- Based in
- France
- Released
- 16 July 2026
- How to get it
- through the Mistral API (mistral-ocr-4-1) and Document AI in Mistral Studio; self-managed deployment for enterprise customers
- Works on
- API, Web, self-hosted
- In the UK
- Yes
- API price
- $4 per 1,000 pages, $5 per 1,000 annotated pages
FAQ
When was Mistral OCR 4.1 released?
Mistral's docs list OCR 4.1 as generally available from 16 July 2026. OCR 4 itself was announced on 23 June 2026.
How much does Mistral OCR 4.1 cost?
$4 per 1,000 pages, or $5 per 1,000 annotated pages.
What is new in Mistral OCR 4?
It returns bounding boxes, block types and confidence scores alongside the text, and supports 170 languages.
Facts from the maker's own pages and the sources listed. Checked 2026-10-11.
Similar tools
Devstral 2
Mistral's open-weight coding model family, now retired from the API in favour of Mistral Medium 3.5
Mistral 3
Mistral's December 2025 family of open models, from the small Ministral 3 sizes up to Mistral Large 3
Mistral Small 4
An open Mistral model that handles chat, reasoning, coding and images in one 119B mixture-of-experts model
Mistral OCR 3
Mistral's December 2025 document reading model that pulls text, tables and images out of PDFs and scans